"So now the question is, will

"So now the question is, will it actually print, and how much work will it take to get a truly usable wheel?  It would be great if it just needed a quick spin in a drill or lathe with a little fine sandpaper to smooth the tread.'

 The shapeways stuff I've used is about as smooth as an unmachined casting so I think hitting the flange and tread with a lathe would be adequate. You could bore the axel center hole then machine the rim to match that. Unmachined cast drivers used to be quite common so you might save some trouble by asking around sites where old steam modelers hang out. I've seen buckets of them in the past that no one wanted once newer locos  were available. ....DaveB 

No lathe here

How bad would the tread surface be for something like this?  How bad would it be if I tried to run it straight out of the box?  I don't think I would do so, but I'm trying to get an idea of just how rough the surface would be.  If it's close, a quick turn over some sandpaper ought to do it, but if heavier machining is needed, that becomes a problem...

If I had a lathe or a way to machine a casting, I wouldn't be looking so hard for a driver in the first place!  I'd just make a suitable center and go from there.  My goal is to come up with something I can do in my garage, using nothing fancier than an electric drill.

Polished Nickel?

For a few dollars more, they offer polished, nickel plated steel as well:

Maybe that would have a smooth enough surface that it wouldn't need any further treatment.  It looks like I may have to place an order in the near future.

"How bad would the tread

"How bad would the tread surface be for something like this?  How bad would it be if I tried to run it straight out of the box?  I don't think I would do so, but I'm trying to get an idea of just how rough the surface would be.  If it's close, a quick turn over some sandpaper ought to do it, but if heavier machining is needed, that becomes a problem..."

    The surface would not be good for power pickup if the metal wheels they cast are the same roughness as the printed resin parts. The surface is similar to a sintered iron wheels that has pitted from arcing if you are familiar with that.    I imagine you could chuck a wheel in a drill and smooth the tread and flange with a file and sandpaper( in fact I've done it on N locos to make the flanges clear code 55 track spikes) but it would be a lot of work compared to buying a Bachman or other close to correct  loco and using the drive :> )

  I'd check ebay before going too far and see if there are any classic diecast  locos that might be drive donors or perhaps some old brass that's close enough. You might even find some separate drivers for sale, I see them occasionally without even looking for them. The other source could be train shows, I see lots of parts at O/S scale west so I assume HO shows would also have some? ....DaveB

Counterweights?

Drawing looks good. The project also.

Real drivers have weight to counterbalance the connection to the rods. Different size weights depending on how many side rods are involved. Does this affect your cost per unit?

DrJolS

Cost per unit

Only slightly.  I am thinking about designing a set of counterweights to be 3D printed as well (probably as part of the axle center print), but it's also pretty easy to make some out of styrene.

I have two distinct goals here.  One is to learn whether this is even a viable option to get an accurate, quality wheel which is not otherwise available (even if I can get the right size, forget getting the right number of spokes and crank throw).  The other is to try to put together the most complete model possible from 3D printed parts.  So far it looks like the body, chassis, drivers, and gears might all be practical.  That leaves only a motor, wiring, couplers, paint and decals.  The cost may be high, but it might be reasonable for a few models which have never been produced.  At least, I intend to try and find out.

Hope I'm not getting too excited too soon

I just designed a driver pair with rounded spokes.  Shapeways says they can print it for $13.70, which I don't think is a bad deal.  Greenway wants $18 for a pair of drivers, and most of the drivers from the UK are at least that high when you add exchange rates, shipping, etc.

I've also contacted Shapeways, asking if they would consider making nickel plated brass, or cast nickel silver available.  Either one would be an ideal material for a driver, in my opinion.

Great job Ken!

I'm also interested.  If you are successful, I might be interested a few 30 and 36 inch drivers.  Is it just me or did you miss a spoke?  Maybe it is the view.  The top center spoke still looks flat.  Keep up the good work and don't worry about getting excited.

Den

If you're going to be

If you're going to be machining them in ANY way, even sandpaper in a drill, you'll want to do the plating yourself or have it done locally after you work them. Otherwise you'll just be taking the plating off when you machine it.

Insulation?

Split chassis

I would print the wheelset as a single piece, and then cut it in half.  I designed a small 1mm diameter center section for just that reason.  Then each half of the axle could be inserted into an insulating center section, much like an Athearn half-axle, except mine would have a square center to force the drivers to be quartered.

I think....

.... you may have to check if the axle ends up exactly centered on the wheel, and if the wheel is exactly round. Just a little deviation will make the loco run real funny. Jose.

Pretty much works...

Hi Ken,

Here are a couple that I printed at Shapeways a few years ago, when we evidently still had pennies in Canada.

As you can see, I lucked out, and they printed them flat.  Otherwise, the tread would be completely unusable.  either way, you would certainly have to machine them as they are too rough to take as they are.  I don't have a lathe either, and so, I gave them to my friend, Andrew, to try out on his lathe.

He reported that it machined quite nicely, but it is very hard to file.  I think it would probably respond okay to a form tool if you had one, and took it slow.  Andrew was less forgiving about the face of the spokes than I am, and resolved to clean them up.  So, they wound up pretty square. 

I also managed to violate the minimum wall thickness, but don't ask me why they allowed it.  Shapeways recently announced an investment casting service, and this might be a good way to go for drivers as you could have thinner spokes.  I don't think you can get away from machining the tread, though.

I shall be very interested to hear how you make out with this experiment.  I need some P87 drivers in the next year to eighteen months if I'm to win my beer from @mesimpson!

Time to try it

Thanks for the photos, Rene.  I'm glad I'm not the only one who has at least considered this method of getting wheels.

I decided to bite the bullet and place an order.  No point thinking about it if I'm not going to try the idea eventually.  I actually ordered three things:

First, the round spoke version which I have every reason to believe will print:

Next, a version with 0.75mm spokes - technically too thin, but I figured I'd try it anyway:

And finally, a set of plastic axle centers to connect the wheels:

We'll see if all the wheels print, and if any of them work without post-processing.  I am hopeful.

57" drivers

Don't know if this would help.

http://www.ebay.com/itm/HO-Vintage-57-Drivers-for-0-6-0-Hallmark-PFM-Sunset-United-PSC-On30-brass-/281421798121?pt=Model_RR_Trains&hash=item41860bdae9

Following this with interest,

Trying something new

@ lvflats:

Those might be the right diameter, but I cannot be sure, and I don't like relying on finding old stock.  The whole point of scratchbuilding is being able to create an accurate model of whatever you need, not make do with whatever you can find.  In that spirit, I want to learn if it is possible to create whatever drivers I need, rather than hoping to find some off of a specific brass locomotive built sometime in the last half century.

@ Paul Schmidt:

Learning CAD just takes time.  It helps if you have the right mentality, so that you can think like the program does and understand what it's doing or expecting to see from you.

As for the frame, I haven't decided.  I've played with a few plastic frames, incorporating springs as part of the design.  If I decide to skip that feature, I might go with a metal frame.  They offer brass and steel, either of which would be suitable for a locomotive frame.  Another option would be to leave enough room elsewhere than a plastic frame is suitable, and the weights are easy to make.  I'll have to evaluate the options when the design progresses.

Make the frame plastic, with

Make the frame plastic with integrated springs... and leave cavities to be filled from the top with tungsten putty, available as Pinewood Derby weight from Woodland Scenics, or moldable fishing sinkers.

The results are in...

I got another box from Shapeways this morning, and my wheels were inside!  Here's what I found:

That's the polished, nickel plated steel.  So clearly the idea of the polishing and plating yielding a usable tread didn't work out.  The wheels do in fact roll, and I suppose you could use them as-is if you had to, but it would be a rough ride!

To give a little sense of scale, that's a 57" driver, RP25, code 88, and the round portion of the stub axle is 1/8" diameter.  While not microscopic, they're not exactly gigantic either, so the surface texture looks a little worse in the photo that it looks (or feels) in person.

I can't say that I'm all that surprised.  It was a stretch to hope that I'd have a perfect wheel first try.  No matter.  I cut the axle in half and chucked one of the wheels into my poor-man's lathe (old drill bolted to a piece of plywood), and attacked the tread with a file.  I tried my homemade cutting tools, but it was slow going without either proper tools or a proper tool holder.  It turns out that at high speed, my miniature files were the best tools for the job, and it cut quite quickly.  I cleaned up the tread, the face, the back, and the flange.  Here's the result, compared to an untouched wheel:

It's still not perfect, but definitely usable at this point.  One interesting thing to note is that the surfaces of the wheel centers are quite different, even though they were in exactly the same orientation in the drawing.  Oh well, no big deal.

I cleaned up the second wheel, then cut one of my axle centers out and pushed it onto the square post on the axle.  The fit is tighter than I expected, but it certainly seems to hold well.  We'll see if it holds up without splitting.  Here's an assembled wheelset, next to an untouched one:

So, here's what I've learned:

• You cannot reliably print a driver which will require no post-processing.  I didn't really think you could, but it was interesting to try.
• You CAN print a usable driver, provided you have a way to clean up the tread.  A drill and a file works.
• Polishing and nickel plating are not worth the expense or delay in this case.
• Self-quartering and self-gauging seems perfectly doable.  After assembly, I checked my wheelset against an NMRA gauge and it's spot on.
• There is no obvious difference between a 1mm thick spoke and a 0.75mm thick spoke.  I have not measured them, but they don't look any different to me.  Judge for yourself - there's one of each in the last photo.
• The most effective way to get a good looking driver with this method will probably be to make a separate plastic center to go into a steel driver, similar to the way many older Bachmann drivers are made.  That way the tread can be cleaned up, and then a much smoother plastic center can be inserted afterward.  Cleaning up these drivers may be a chore, if it's even practical at all.

Yet to be determined:

• I don't know how well the axles will fit any kind of bearings.  I am hopeful that a piece of brass tubing will work as a bearing, and maybe I will not have to dress the surface of the printed axle.
• I do not know if the axle centers will survive repeated assembly and disassembly.  They may stretch or wear enough to only work once.  I printed extra for a reason.
• Ido know know if the plastic centers will tend to split.  I am hopeful that they will not, as there's no obvious line of weakness for them to split along, as with an injection molded center or gear.
• It will be interesting to see how these wheels look once they have counterweights installed and they are painted.  I hope that the rough surface will not be as noticable.

"To give a little sense of

"To give a little sense of scale, that's a 57" driver, RP25, code 88, and the round portion of the stub axle is 1/8" diameter.  While not microscopic, they're not exactly gigantic either, so the surface texture looks a little worse in the photo that it looks (or feels) in person."

  That's similar to the way the old sand cast drivers came out of the mold. If one had a lathe and knew a bit about machining they would, probably clean up and look fine for the job. Until 3D printers get better resolution I don't see an easy way to print drivers. If you could find pre-machined tires I guess you could press them onto plastic centers and do something similar for the axle bearings? Nice to see someone experimenting with this stuff, thanks for posting the photos.....DaveB

Tolerable quality?

I don't know, I think these will work and look decent.  I've got another test print on the way - a frame and gears this time.  I'm intentionally pushing the limits of what is practical with 3D printing, just to see what works.

One of the nice things about these, compared to most of the old cast drivers, is that they are steel.  That means they're the right color, even after machining.  No plating to wear off, either.  I need to check and see if they're magnetic, but I don't think it would a bad thing even if they are.

Interesting

I believe nickel is non-magnetic. Also a quick check shows me that the process used by 3D printers for nickel is a sintered process. Same as what was used on some of the early model wheels. This should give your DCC fits as those wheels pick up dirt in those pits. The best wheel will be a turned stainless steel on some kind of center, be it plastic injected or 3D printed. I don't think your going to get away from machining at lest a tire for your wheel. And from what I'm seeing of the finish I have my doubts as to whether  it will work for bearings or gearing. Just my analysis from what I've seen so far. Wish you luck though on the next print.

Bernd 

More experimentation is needed

Shapeways says that the process is to laser sinter stainless steel powder into the desired shape, and then to infuse the result with bronze.  For the nickel plated parts, they then polish and electroplate them.  So these wheels are a combination of stainless steel, bronze, and nickel, with the nickel being strictly cosmetic (and honestly a waste in this case).  I had some concerns about turning the tread and revealing a bronze-colored surface, but it is a very nice silver color.  I also just checked, and the wheels are not appreciably magnetic.

Yeah, I've wondered about that.  I'm thinking about trying to work a little molten solder into the surface as a filler.  Maybe get the wheel turning, aim a torch at the tread, and apply solder to the (slowly) turning wheel, then wipe it down.  After cooling, re-machine the surface lightly to remove any excess solder.

Or maybe I'll just go battery powered!

I think that gearing is out of the question in this material, at least in HO scale and probably in any of the indoor model railroad scales.  I also do not think that bearings will be a good idea, though I am interested in other materials.  The 3D printed axle running in a brass bearing, on the other hand, I am curious about.  Here's my theory:

The surface is quite rough, but it is reasonably round, and the roughness is not sharp.  Therefore, if I put the axle into something like a brass bearing, the high points should make contact and run smoothly.  The fact that not all of the axle is making contact with the bearing might reduce friction somewhat, or it might simply accelerate wear.  Maybe it will alow oil to stay between the surfaces without being squeezed out, too.

I may find that I have to turn the axle down slightly, if the combination of roughness and plating have made it larger than the i.d. of the tube I plan to use as a bearing.  If so, then I suspect I can make a surface good enough to last a reasonable time without too much trouble.

Very cool

Nice work, Ken.  I especially like the idea of the printed axle centers.

What do you think of spraying the centers with filling primer and seeing how long it takes to sand back to a smoothish finish?

I would think the rough axles are going to give problems when they wear.  The result would be a bunch of wiggle, and finally binding with the connecting rods.

Cheers

Rene